1. Field of the Invention
The present invention is directed to a method and system for providing rich media content over a computer network and more particularly to a highly reliable, transparent process for displaying high-quality online advertising imagery.
2. Related Art
Computer networks, including the Internet, are a new and rapidly growing means of communication. There are currently over 300 million network users worldwide, and that number is expected to double in less than five years according to the Computer Industry Almanac (www.c-i-a.com). Because of their numbers, and because they are believed to be high-end consumers, users of the Internet and other computer networks are an attractive audience for advertising messages. According to figures from the Internet Advertising Bureau (“IAB”; www.iab.net), expenditures for Internet advertising are growing even faster than the number of network users—at almost 25 percent per quarter—and currently total over $7 billion per year.
The current standard for Internet advertising is the banner ad, a cartoon-like color image that occupies a fixed part of a web page. Banner ads usually come in one of a small number of standard sizes, and they sometimes include crude animation created by rapidly and repeatedly superposing a small number of images in the same space. Banner ads comprise the majority of Internet advertising—historically over half of total expenditures and around three-quarters of that on discrete ads (as opposed to sponsorships) according to the IAB.
The content of a banner ad image is contained in a computer file that is interpreted and displayed by a network user's web-browser program (e.g., Netscape Navigator, Microsoft Internet Explorer). Almost all banner ad files use the GIF89a format, which can be interpreted and displayed by all major web-browser programs currently on the market.
Each banner ad's file, plus the file containing the content of the host web page itself, must be transmitted to, and stored on, a network user's computer before the complete web page (ads and content) can be displayed. There may be many—sometimes dozens—of banner ads on a page, and advertisers often demand that the page be configured to display the ads first. To keep network users from waiting too long to see the content of their pages, web page owners frequently impose limits on the size of the banner ad files they will accept. These limits, in turn, sharply constrain the appearance of banner ads, e.g., by reducing the number of images per ad or the number of items per image, or by restricting the variety of colors or level of detail in an item or image.
Such limitations have combined to reduce the effectiveness of banner ads in two ways. First, they are—and because of file size limitations must remain—crude. Because banner ads are typically simple cartoons in an environment of increasingly rich and complex media, the viewing audience is becoming less responsive to the ads. The average “click-through” rate, the rate at which viewers respond to a banner ad by “clicking” on it and being transferred to the web site advertised by the banner, is falling rapidly—by almost an order of magnitude during the past five years, according to contemporary estimates in Advertising Age magazine. Secondly, despite limitations on their file sizes, banner ads often delay viewing of web page content to the point that users routinely redirect their browsers away in frustration; the very presence of the ads lowers their own viewership. Dwell-time, a measure of the time an average network user spends looking at a single page, is also declining.
In brief, banner ads, though they dominate advertising on the Internet and are so much the standard that they are directly interpretable by every major web-browser program, are an obsolete and increasingly self-defeating technology.
Advertisers, aware of the limitations of banner ads, have tried two approaches to improve upon them. The first approach is to replace the cartoon-like banner ad images with video ads, i.e., online ads that use moving, photographic-quality images rather than simple single images or series of a few simple single images. The second approach, referred to generally as interstitial ads, sometimes called pop-up ads, avoids some of the technical problems of banner ads and video ads by effectively separating the interstitial ads from their host web pages. Both approaches, though, like the banner ad approach, have run into technical and consumer-response barriers.
Video ads, unlike GIF89a-format images, cannot be displayed directly by web-browser programs. Instead, they are encoded in one of several specialized formats (e.g., MPEG, QTF, AVI) and are displayed by separate—and similarly specialized—video replay programs. Such video replay programs are separate from the web-browser programs, but they are compatible with the browsers and sometimes are referred to as plug-ins to the browsers. Some replay programs are distributed with computer operating systems, while others are available separately, either for free or at a cost. Some replay programs can interpret more than one format, although none can interpret even a large minority of the existing variety of formats, and file formats are different enough that any multi-format program is essentially a package of single-format programs. Because of the multiplicity of formats and distribution methods, and because many formats are in fairly wide use, it is unlikely that any single standard will emerge in the near future. In other words, unlike GIF image technology, the technology of computer video replay is non-standardized and is functionally complicated, and it is likely to remain so.
To view a video, a user's computer must receive at least part of the file and then activate and run a compatible replay program. This process frequently includes one or more “dialogues” between the computer and the user. For example, if the file is to be stored and then replayed, the user must specify a storage name (or approve the computer's choice) and then activate the replay. If the computer cannot determine which replay program to use, the user must specify it. And if there is no compatible replay program on the computer, the user must either cancel the replay or spend time (and possibly money) identifying, locating, and obtaining one. The complexity and time requirements of this process can be daunting even if the user actively seeks to view the video; for advertising, which needs to be entirely passive and nearly instantaneous, these are major barriers.
Another barrier to the use of video ads is imposed by the sheer size of most video files. Video images, like any other computer data, are stored and transmitted as computer files. Because of the complexity of the images (color, resolution, etc.) and the number of images in a file (usually thousands), video files typically are very large. For example, a word-processing document file may be on the order of a few dozen kilobytes (KB), and a typical web page file may be around a hundred KB, but even a thirty-second video will require a file size of thousands of KB (i.e., several megabytes, or MB). Using a standard telephone modem connection, whose transmission rate is limited by telephone technology and federal regulation, not by computer modem technology, a file of even a few megabytes can require many minutes to receive and store. For example, with a 56K modem and an effective transmission rate of over 50K, transmission requires at least three minutes per megabyte. Faster connections (e.g., via DSL or institutional intra-net) reduce this time substantially, but not so much that the file transmission does not cause a noticeable interruption. Further, those types of connections are available to only a portion of the user population. As noted above, computer users are becoming increasingly impatient with any delay, especially for the sake of advertising, and in any case most advertisers are unwilling to limit their messages to only a portion of the population.
To address the problem of file size, specialized protocols were developed that allow near-real-time playback. Sometimes called streaming, these protocols begin playback when only a portion of the file has been transmitted and stored on the user's computer (i.e., buffered). Later portions of the file are transmitted as earlier ones play back, and the parameters of the process are calibrated so that, if transmission is not interrupted, playback is continuous. The practical flaw in this approach is that transmission, particularly of large files, frequently is interrupted. Net congestion, transmission errors requiring retransmission, competing demands on the transmitting computer, and other causes, can interrupt the transmission flow long enough that the buffer is completely played out, and then playback stops until enough new data have been received. The effect on the user is that streaming video (or audio) either is occasionally interrupted by long pauses or has a jerky quality caused by frequent micro-pauses (the former with a large buffer size and the latter with a small one). These types of interruptions are unacceptable to advertisers, whose imagery requires seamless replay. Further, streaming video is subject to the same requirements as non-streaming video for identifying, obtaining, and/or activating a compatible replay program.
U.S. Pat. No. 6,029,200 to Beckerman et al. discloses a process that uses streaming video and provides a more automated approach to selecting a replay program. The “client” computer (e.g., the network user's computer) is offered a list of multiple versions of a video file, each in a different format, in a predetermined order, until a compatible format—if any—is found. This makes it more likely that the video eventually can be viewed by the user, but it requires the user's computer to know how to interpret and choose among the list of “offers,” which presumably requires specialized software. It is not clear how noticeable the process would be to the user (e.g., delay, “dialogues”), and because it applies only to streaming video, the process does not address the problem of interruptions in playback. It thus can be considered another format, albeit a somewhat generalized one, but with questionable application to advertising.
Processes developed more recently by bluestreak.com, Inc. (www.bluestreak.com) and AudioBase, Inc. (www.audiobase.com) also use streaming, and they circumvent the replay compatibility issue by transmitting their own replay programs along with the data files. These are examples of the usage of “push” technology, as used, e.g., in U.S. Pat. No. 5,740,549, issued Apr. 14, 1998 to Reilly et al. These processes are reasonably rapid—although still noticeable to the user—because neither handles full video; bluestreak offers audio, GWF-like animation, and other cartoon-like special effects, and AudioBase handles strictly audio. Both also are subject to the problems of streaming, such as “stuttering” and interruptions in playback. They thus are suitable for certain narrow, specialized forms of advertising, but their transmission delays are still non-negligible, and like other uses of streaming, their unreliable replay renders them problematic for advertising.
In summary, attempts to date to put video advertising onto Internet web pages have largely failed because of two fundamental technical characteristics of computer video—lack of standardization and very large file size—and their implications. Computer users are generally unwilling either to wait for large files to be transmitted or to take active steps to ensure a smooth replay, especially for the sake of viewing an advertisement. Advertisers are unwilling to spend money and effort on technologies that cannot reliably deliver uninterrupted imagery to a wide audience. What would satisfy both users and advertisers, but is lacking in the prior art, is a means for reliably delivering video ads without any interruption of the user's viewing experience. As a consequence, video advertising has been and remains a small and static fraction of all Internet advertising; expenditures on video have consistently been just a few percent of the total.
Interstitial advertisements—sometimes called pop-up ads—bypass some of the technical problems of on-page banner ads and video ads by effectively separating the online ads from their host web pages. The content of an interstitial ad is transmitted separately from those of its host web page. Transmission begins immediately after the host web page has been fully transmitted and while it is being displayed (i.e., during the “interstices” between other web page transmissions), and it may continue once the ad itself has begun displaying. The ad is displayed in a new “window” or other dedicated display area, either immediately after the host page is fully displayed (thus “popping up” in front of the host page) or when the user signals that he/she has finished reading the host page by closing it or activating a link to a new page. Interstitial ads can include GIF images, video, audio, or any other web page elements; they are essentially specialized web pages comprised entirely of advertising. Because they are transmitted separately, they do not delay the display of the host web page, and because the user presumably is occupied for some time reading the host web page, the ads can take much longer to transmit than on-page ads without seriously annoying the user.
Interstitial transmission of advertising is taught in U.S. Pat. No. 5,305,195, issued Apr. 19, 1994 to Murphy, for use on specialized computer terminals such as bank ATMs and school registration stations. More recent applications include U.S. Pat. No. 5,913,040, issued Jun. 15, 1999 to Rakavy et al.; U.S. Pat. No. 5,737,619, issued Apr. 7, 1998 and U.S. Pat. No. 5,572,643, issued Nov. 5, 1996, to Judson; and U.S. Pat. No. 5,604,542, issued Feb. 18, 1997 to Dedrick.
The principal problem with interstitial ads is that, as dwell time statistics show, users' patience still limits the time available for transmission. While a user is reading the host page, there typically is sufficient download time for banner ads, other GWF images, other static images, simple animations, a streaming video buffer, and usually audio and other animated elements and associated programs such as those produced by the bluestreak and AudioBase processes. However, there is not sufficient time to transmit more than one buffer worth of a video file, and there is no opportunity to create a “dialogue” with the user until the interstitial ad is displayed. Consequently, both the number and the type of elements in an interstitial ad are constrained. Additionally, interstital video ads are also subject to the same problems of non-standardization and display reliability as with on-page video ads.
A process developed by Unicast Communications Corp., disclosed in International Publication No. WO 99/60504, published Nov. 25, 1999, partially addresses these problems by installing a program on the user's computer that ensures that transmission of ads (and accompanying playback programs if any) is as “polite” as possible. The concept of “polite” transmission—i.e., file transmission minimally noticeable to the user—avoids some of the problems associated with streaming by fully storing (or caching) ads on the user's computer before displaying them. The Unicast program also checks that any necessary playback programs are available on the user's computer before the ad is displayed. However, such a process is not practically applicable to video ads, because large files take a long time to transmit no matter how politely, and video replay programs are even larger than video ads and are seldom available for transmission along with them. It is also as yet unclear whether such a powerful program will be fully compatible with most network users' operating systems and browser programs, or whether privacy concerns will limit its functionality or popularity.
Overalls while interstitial ads solve some of the problems of computer network advertising, they as yet have shown only partial success at dealing with the non-standardization and file size issues associated with video ads, or with file sizes in general. Attempts to create a “smart” solution have only added new problems. At least partially as a result, interstitials historically have accounted for less than a tenth of Internet advertising expenditures, and that share has shrunk by almost half in the last year according to the IAB.
In summary, what is needed but is missing in the prior art, is a highly reliable, entirely transparent process for displaying high-quality rich media content over a computer network.